At risk of being left behind on AI, SMBs focus on the art of the possible

When companies like Google and Microsoft can afford to spend hundreds of millions of dollars on AI, some IT experts have questioned how SMBs can tap into the revolution.

Many AI projects have huge upfront costs — up to $200,000 for coding assistants, $1 million to embed generative AI in custom apps, $6.5 million to fine-tune gen AI models, and $20 million to build custom models from scratch, according to recent estimates from Gartner. Those costs don’t include recurring costs, which can run into the thousands of dollars per user each year.

At the same time, many company executives believe AI will widen the gap between the haves and have-nots, according to a survey of more than 2,800 director- to C-suite-level respondents across 16 countries by the Deloitte AI Institute.

More than half of the respondents expected that widespread use of gen AI will centralize power in the global economy and increase economic inequality, according to the survey, released in January. Just one in five respondents believed gen AI would decrease economic inequality.

The barrier to entry

The big issue for SMBs is the cost of the computing power and related expenses needed to run modern AI models, says Tony Fernandes, CEO and chief AI officer at HumanFocused.AI, an AI consulting firm.

“Does anyone believe that big tech is just going to eat this expense?” he says. “Services using all this upgraded hardware are going to be more expensive than previous generations, yet most businesses are not taking that reality into account. SMBs are particularly vulnerable to these cost increases.”

Most SMBs don’t have the resources to create and maintain their own AI models, and they will need to work with partners to run AI models, he adds.

“They will be handing over customer data to AI companies that reserve the right to use it for their own purposes,” Fernandes says. “There is a lot of legal peril that SMBs have not taken into account.”

In addition, many companies are still looking for the ROI in their AI projects, and SMBs hoping to reduce costs and cut headcount may instead need to hire prompt engineers to get value out of their investments, he adds.

“Overall, there is more hype than reality in terms of the benefits of AI, and SMBs should let things play out a bit more before they make an investment,” he adds. “It is too risky, and its ROI is unproven.”

Build it with a partner

Kristopher Stuart, COO of Bloomin Blinds, has a much different view. The window treatment company, with 17 direct employees and franchises in 35 states, is now beta testing a small language model created with Revscale AI.

The BloomScale AI, owned by Bloomin Blinds and scheduled to fully launch in early 2025, automates inbound and outbound sales, with an AI-powered call center. It also automates some marketing functions, purchasing, and KPI management.

The first phase of the AI development took less than three months. The cost, Stuart says, was less than $1 million. That’s a small fraction of the price the company was quoted when it was last looking for a standalone CRM.

As it looked for partner to help with its project, Bloomin Blinds saw many AI developers looking for new use cases, Stuart says. “There’s so many developers trying to find niches in the market,” he adds.  “We had a very clear idea of what we were looking for, and they were able to take that and run with it.”

BloomScale AI, through the AI-driven call center, will allow the company to retrain its seven-member inbound customer support team to focus on sales, creating more revenue opportunities.

“Our franchisees just got a no-charge seven-person inside sales team added to the arsenal,” Stuart says.

Bloomin Blinds is continually looking for new ways to automate processes to cut costs while providing customers with better service, he adds. At some point soon, robots like the Tesla Optimus 2 may be able to install window blinds in customer houses with pinpoint precision, he suggests.

A constellation of AIs

AI-as-a-service may be another model for SMBs, says Matthew Marolda, chief innovation officer at Acrisure, a large insurance broker and financial services company. Acrisure is using AI technologies in several ways, including matching potential clients with insurance carriers, searching for potential new customers, and helping employees find experts within its 17,000-strong workforce.

Marolda sees an AI market developing where vendors create tools for specific small-business needs, such as accounting. Each SMB could potentially use a half dozen different low-cost AI tools, based on processes they want to automate.

He expects some large companies to offer their internal AI tools to smaller businesses. One of Acrisure’s goals with its AI initiatives is to help its small-business customers find matches for their insurance needs, he says.

He sees the video streaming market as an apt analogy for how an AI SMB market might evolve. Dozens of video streaming services have sprung up, some with a more general focus and some “super esoteric.”

“You’re almost shifting to a new paradigm, and I wouldn’t be surprised if tools that are AI-driven for SMB kind of follow similar paths,” he adds. “The first handful are really powerful, and next thing, there’s another one and another one and another one. And you select from this constellation of tools.”

Behind the scenes

While AI can have many uses, the sweet spot for SMBs will be in automating back-office functions that cost significant employee time or that SMBs can’t provide themselves, says Erik Severinghaus, CEO of Bloomfilter, vendor of AI tools to monitor software development.

While chatbots and digital assistants can provide value to SMBs, they can also use AI for data analysis, inventory management, personalized marketing, and fraud detection.

Process automation, such as automating bookkeeping and payroll jobs or managing social media accounts and email marketing campaigns, can bring huge benefits to small businesses, he says.

AI-based process automation tools will make SMB employees more efficient and effective, Severinghaus believes, ultimately impacting the bottom line. Owners and CEOs of SMBs often get caught up on driving the business forward, while ignoring ways to improve the business from within, he says.

“What AI, really for the first time, gives us the opportunity to do is apply flexible technology to do process improvement, to do efficiency analysis,” he says. “That’s where I think it’s particularly disruptive and particularly interesting compared to other technology that’s that sort of come before it.”

A recent survey from BILL, a financial operations platform for SMBs, found that 85% of SMBs are enthusiastic about using AI. Still, Bloomin Blinds’ Stuart sees some resistance to AI, driven by a misunderstanding and fear of the technology. SMBs need to get over those concerns or risk being left behind, he says.

The concern “really revolves around that control humans demand on things they don’t understand,” he adds. “This is truly one of those evolutionary moments of humanity where we have to change our thinking because we are no longer the top of the food chain.”

The new haves and have nots are going to be business that embraced AI and those that didn’t, he adds. “With the pace of business, the pace of advancement, if you’re not on the train, it’s just simply distancing itself from every single day.”